AMN helping operators connect 108m new subscribers in rural Africa

Africa Mobile Networks (AMN) has an interesting business model. They target similar opportunities to conventional independent towercos, only with the key parameters reversed! Where towercos provide passive infrastructure only to tier one MNO anchor clients needing extra capacity in high value, dense urban areas of established countries, AMN targets either tier one or tier two anchor tenants with a proposition that combines active and passive infrastructure and which focuses on rural areas of countries with less mature coverage.

TowerXchange: Where does AMN fit in the tower industry ecosystem?

AMN is a towerco with a difference!

AMN is essentially a tower company but managing the active as well as passive components on a turnkey basis and taking a share of the risk on the revenue side. AMN handles O&M and traffic management - we even handle airtime distribution! We deliver traffic to the operator’s switch; they don’t have to deploy any capex.

AMN provides an access service to connect an operator to their own subscribers, enabling operators to increase their penetration deeper into rural communities.

TowerXchange: How would you describe the market segment AMN is addressing?

About two thirds of the population in Sub Saharan Africa (SSA) are served by the mobile industry today. But 300mn Africans have no service whatsoever. In the next five years, 200mn of that remaining 300mn population will be served - representing a US$4bn revenue annual market. We divide those 300mn people into three groups:

• 100m who are likely to be served by existing MNOs through normal expansion over the next two years as they push from high density urban areas (which are increasingly saturated and highly competitive) into rural areas in search of growth

•  Around 73mn people who nobody, including AMN, can serve commercially - there are just too few people spread across too wide an area. These people should be the focus for the investment of Universal Access or Universal Service Funds, whose objective should be to build infrastructure where nobody else ever will

•  AMN is interested in a middle segment of a little over 100mn unserved people, where we’re not competing with existing operators today. This segment is not likely to be served anytime soon, but there is still sufficient population density to serve them profitably under our business model.

TowerXchange: What has been your progress to date in your first country, Benin?

We aim to build 200 towers in Benin.

We’ve built up some good momentum and are aiming to have our first site in Benin commissioned in the next couple of months, achieving our objective to become a revenue generating company by the end of the year.

Our first site is in Koabagou in northern Benin, over 500km north of Cotonou. The main settlement has a population close to 4,000, then the site will also cover three other communities with populations of 500-1,000, meaning we’re giving access to 6-7,000 people, well in excess of our minimum criteria. This pilot site will give us key operational data, which we’ll use to update our models and inform our full rollout.

TowerXchange: What are AMN’s minimum criteria for site selection?

We’re interested in sites serving a population of over 4,000, with no existing service whatsoever from operators. That generates enough traffic and enough revenue to make our business model work.

TowerXchange: Low capex and, in particular, opex are critical to your business model - tell us about your procurement process.

Minimising both capex and opex are indeed critical - over a number of years opex becomes the dominant component of TCO.

We have a very thorough procurement process, broken down into fundamental components: towers, communications equipment, solar power systems, VSAT and satellite capacity, which is the biggest single item in our opex budget.

TowerXchange kindly introduced us to a good number of very solid potential partners, and we had over 30 proposals in response to our multiple RFPs. We do our due diligence to ensure we’re getting the best value and to ensure we choose partners we can rely on.

We selected Leadcom for the turnkey build work of the first “proof of concept” base station in Benin. Once they grasped the project, Leadcom decided they wanted to be part of it - they’ll do the installations, civil works, site surveys and O&M. Leadcom already had people on the ground in Benin, and they’re already shipping the tower and solar EPS.

Altobridge has agreed to ship a BSC and BTS to Benin together with their engineer to start core network integration.

It was interesting to see the differences between proposals from several different tower manufacturers. Some quotes at the bottom end were remarkably close to each other, while we received other bids that were double the cost of the cheapest! We gave candid feedback, without giving away anything confidential, but as a buyer I’m not going to waste my time negotiating with the most expensive bidder! From our experience in Ghana, we know what the market price levels are; once you know you’re at right level on price, it’s important to select right partner, not just the cheapest.

TowerXchange: What have you found to be the more economical approach, buying and integrating best of breed solutions from each component category, or buying an end to end turnkey solution?

When we sent out our RFP, we were contacted by a number of suppliers suggesting they deliver an end to end service. We encouraged them to take any multiple components of our RFPs, and to send their proposal. Generally the end to end proposals were not attractive. There’s always a tradeoff between the margin a consolidated turnkey supplier would have to add, versus buying equipment direct. We thought that their economies of scale might offset their margins, but ultimately the end to end service providers proposals have turned out to be quite disappointing.

So we reverted to Plan A; to do things ourselves!

TowerXchange: What is deployed at a standard AMN cell site, and what does it cost?

Michael Darcy, CEO, AMN: We undertake a huge amount of analysis to optimise our approach for each cell site. While a conventional cell site might cost US$200-250k, including active components, we can reduce our capex to less than one third of that figure and our TCO to 20-25% by deploying a solution optimised for rural markets. We have a large, low power cell that runs on solar power and batteries, so there’s no diesel logistics, which minimises truck rolls. Our MTBF is 2.5-3 years.

We deploy a two TRX, 2G site, with a lower cost base that means a lower population can be served economically, but can expand in a very modular fashion to add 2G capacity as required and - if demand justifies - also 3G functionality.

TowerXchange: How has AMN overcome the transmission challenges of connecting remote rural settlements?

When I was previously deploying rural telephony for one of Africa’s leading MNOs, they couldn’t economically connect to settlements as large as 14,000 population because they would have needed three repeaters just to get the traffic back, and repeater stations cost almost the same as a BST since diesel is the biggest cost. AMN’s sites don’t burn any diesel, and we use a VSAT link so there aren’t the transmission constraints you get with microwave backhaul.

For this first site we’ve built a point to point satellite link with a 2m dish on top of a building in Cotonou, and a direct link to the switch from there. Our plan will be to build a mini teleport with a 4.5m dish to connect to remote terminals on all our planned 200 sites. We are currently evaluating proposals from Hughes, iDirect and Gilat for the mini teleport in Cotonou. Both Hughes and iDirect have included loan of equipment for the p2p link for the first link. We’re also evaluating VSAT bids - SES and Eutelsat have both submitted a proposal which includes six months free of charge capacity for the first site.

TowerXchange: How is the network designed in terms of equipment onsite in each country and your Global Network Operations Centre (GNOC)?

Our network is designed to have a local gateway in close proximity to the MNO in each market. We build it, set it up, lock the door and monitor and manage it remotely from our GNOC in Milton Keynes, with local technical support, provided by a company called CBC in Benin.

I’ve relocated to Milton Keynes, from where we’ll be managing all our networks. The GNOC will also oversee traffic management, looking for anomalies, growth and congestion. Our philosophy is to minimise duplication of expertise by centralising these capabilities.

TowerXchange: AMN also get involved in airtime distribution - tell us how that works.

When we were rolling out rural base stations with K-NET in Ghana, sometimes the infrastructure network would be extended faster than the retail network, but we found that even if you don’t make available handsets, SIMs and airtime, people will find a way!

However, making airtime easily accessible will boost revenues and, since our model is based on a revenue share, it’s in our interest to ensure there’s a good distribution channel in rural locations. So we reached agreement with our operator partner in Benin that we’ll become an authorised airtime distributer. Like most of Africa, Benin is dependent on pre-pay scratchcards, which are subject to costs of production, physical distribution and retail margins. So we’re in the process of procuring an e-commerce platform for mobile topup, to be electronically integrated with the operator and to one or more local banks.

Cash management is crucial in Africa, so we’ve developed robust business process to ensure airtime distribution and cash is managed securely. We’ll recruit three or four independent dealers in each settlement connected by our base stations. Each dealer is given unique retailer ID, which they can use to pay in to an AMN bank account at our local bank partner. Their account gets accredited with whatever they play in, plus their margin, then they use an app on a smart phone to distribute topups. This ensures a plentiful supply of airtime, and facilitates cash management, without having to send people round on scooters to collect cash.

Our margins are around 5% after we’ve paid dealers - but the margin this isn’t the reason for getting into the retail business; it’s about ensuring airtime is available, and becoming a net payer to the operator at the end of the month, which puts is in a stronger position in our relationship with the operator.

TowerXchange: So how does AMN make money?

We work on a 50-50 revenue share basis.

Operators love that we’ll fund and build the towers - the revenue share tends to be a longer discussion! Initially we got a lot of “how about 70-30” counterproposals, but I know at 30% I can’t even borrow the money. After dealer margins of 5-15%, depending on incentives and promotions, our operator partners are typically left with about 15% operating profit, and we are at a similar level.

AMN unlocks incremental revenues. To the operator this is ‘found money’ - connecting areas with populations of 4,000 people is not in their business plan. I’ve tried selling rural sites to operators and towercos in my previous roles, but their capex is drawn toward higher yielding investment in additional capacity in dense urban areas, rather than in connecting rural villages. So AMN offers a no-risk, zero capex, shared revenue model which offers a guaranteed positive EBITDA contribution.

TowerXchange: You have a very different model from conventional towercos, who tend to prefer to work with tier one MNOs as the most credit worthy anchor tenants. How does your different business model affect AMN’s preferences for which operators you work with?

We find that our proposition is most compelling to the number two or three operator in a market. For us, the credit worthiness of the partner operator is less important than it is for the towerco since we are willing to manage the airtime distributions and cash collections up front; all we need is a good partner with a license and a switch to put the traffic through.

In Benin, our proposition to our client is to access the 200 base stations we’re building in rural areas, and attract around 700,000 additional subscribers, strengthening and consolidating their competitive position.

AMN are also interested in working with the multinational, tier one MNOs, but frankly we’ve found it easier and faster to build relationships with smaller entities with local decision making authority.

TowerXchange: Does your model lead to significant counterparty risk?

Not really - there’s no reason we can’t be connected to more than one operator in any given market. Once we have our BSC connected to different operators’ switches, we can connect a BST to an additional operator at touch of a button, and we’ll already have the airtime distribution network.

TowerXchange: How does AMN create capital value? And do you have an exit strategy?

Investors always need an exit strategy!

Our business plans show a five-year model making a couple of hundred million dollar EBITDA, so creating shareholder value of over a billion dollars.

While AMN are creating infrastructure that generates revenue and margin - the kind of solid financial investment that might be of interest to a towerco or financial investor, my preference would be an IPO. The AMN topco is a UK registered company and, once our networks are rolled out and generating revenue, I’d like to do an AIM listing - giving an option for a partial or full exit to our investors. Another option might be local IPOs in Africa for some of our operating companies.

TowerXchange: What can you tell us about the financing of AMN?

AMN already has access to significant capital, and we plan to raise US$375mn.

AMN is currently funded with seed capital from more than one private investor, enabling us to progress with stage one of our business plan; business development, engagement with operators, establishing our local entity (AMN Benin Ltd), building the proof of concept site in Benin and bringing it into commercial service.

The investment required to build a network of 200 towers is a different order of magnitude growth investment, and we’ll be seeking equity and institutional investors. We’ll also apply to the debt funds established by DFIs - funds which exist solely for the purpose of infrastructure investment in emerging markets. That debt funding is at competitive rates and can be secured over longer term periods than normal bank funds. Of course we can’t secure growth funding until we’ve got a local OpCo in commercial service and revenue generating - for the DFIs it has got to be a growth investment not a greenfield investment. But if we can prove the concept with one tower generating US$7,000 pcm, we can demonstrate that building ten means we can generate US$70,000 pcm.

TowerXchange: The experience of the management team is obviously critical to investors - please introduce our readers to yourself and your team.

I have 20 years experience in mobile and satellite communications, including nearly 10 years as CEO of the EMEA business of Hughes Communications, where I doubled revenues to over US$100mn. Prior to this I was responsible for the design, development and launch of digital communication satellites, and also for managing multiple large R&D projects for 2G mobile communication systems over satellite. More recently I served as the CEO of K-NET in Ghana, serving West and Central Africa, where we developed a model for rural telephony and deployed base stations for major MNOs. I’m a UK national with an engineering degree and an MBA.

Simon Watts, our CTO, was my Chief Engineer at Hughes Europe for 10 years - he’s one of the world’s foremost experts in VSAT technology. Simon developed the architecture for the Camelot network which runs the UK lottery and connects to teleports in Italy and Germany, as well as large-scale wide area networks for many tens of blue-chip corporate clients across the EMEA region. Simon is now refining the architecture for our total SSA network.

Terry Reynolds is our Chief Operations Officer. Terry managed the Camelot delivery with me at Hughes, installing 28,000 sites in ten months.

We also have a growing team based in Africa, including our VP Business Development Jules Dégila, who is a dual Benin and Canadian national, and got his PhD in Canada. Jules won the contract for us in Benin, and has opportunities lined up in Togo and Cote d’Ivoire, which we’ll address after executing in Benin. Jules does more than business development - he manages local site acquisition, negotiates land lease contracts, and arranges local support partners.

Jules is joined by Emmanuel Pobee on the ground in Africa. Emmanuel is a Ghanaian national with a degree in IT and an MBA. Now AMN’s VP Engineering, he also worked with me at K-NET in Ghana where he was responsible for implementation and operation of rural base stations for Tigo and MTN.

TowerXchange: What is your long-term vision for the expansion for AMN?

We plan to replicate what we’re doing in Benin at least tenfold. We’re targeting a presence in a dozen African countries - we’ll respond opportunistically as markets evolve.

• Population: 9.9mn

• Penetration: 85-90%, depending on source

• GDP per capita (PPP): US$1,700 (CIA Factbook)